6Diaphragm
6.1Introduction
Today, a literature review of studies relating to congenital diaphragmatic hernia (CDH) reveals more than 5700 references. The sheer volume of research in this area reflects ongoing and tireless efforts to unravel the still enigmatic etiology of CDH, and to improve the pre- and postpartum treatment and surgical techniques utilized to treat this condition. The interdisciplinary group from Rotterdam, namely Dick Tibboel, René Wijnen and Hanneke IJsselstijn accepted the challenge of compiling an up-to-date, concise overview of contemporary research approaches and findings. The overlap between CDH and congenital diaphragmatic event ration (CDE) might be of minor relevance. However, the surgical procedure is different, and the French colleagues Anne Schneider, Francesca Borruto and Francois Becmeur discuss this topic in depth.
Dick Tibboel, René Wijnen, Hanneke IJsselstijn
6.2Congenital diaphragmatic hernia (CDH)
6.2.1Introduction
For a number of decades the overall survival rates for newborns with congenital diaphragmatic hernia (CDH) remained around 50%. A number of landmark publications have changed the landscape of the individual patient with CDH and have resulted in a higher level of agreement regarding prenatal risk factors as well as standardization of postnatal therapy [1–9]. In this respect CDH has changed from a surgical emergency for a patient in which the main focus of therapy is directed towards optimal ventilatory support and prevention of longterm pulmonary sequalae. Optimal therapy of pulmonary hypertension, which is due to the abnormal pulmonary vascular morphology, and our lack of knowledge concerning optimal medical therapy are today the main determinants of outcomes in individual patients. To this effect important international initiatives were raised such as the institution of the CDH registry in 1995 [7] as well as the establishment of the CDH EURO Consortium in 2006. More recently a Japanese working group on CDH has been established. Moreover international consensus is available now to determine the way of reporting for patients with CDH [10].
6.2.2Epidemiology of CDH
CDH has a worldwide incidence of 1 in 2500–3000 live births. Based on international collaboration a number of candidate genes as well as environmental factors have been identified as playing a role in the occurrence of CDH in individual offspring. Most of the genes, although located at different chromosomes, are known to have strong relationships with the retinoic acid pathway [11–13]. In the most used animal model of CDH (the nitrofen rodent model) the crucial step is interference with the enzyme RALDH2 as a pivotal step in retinoic acid metabolism, at least in experimental models. CDH may occur as an isolated anomaly but the increasing use of prenatal ultrasound and MRI has revealed a significant number of fetuses with associated anomalies (cardiac, renal) while modern genetic techniques have increased the yield of genetic abnormalities (array and next-generation sequencing).
6.2.3Pathophysiology of CDH
For a long time, the hypoplastic lungs of CDH patients have been considered as immature and resembling the lungs of preterm born infants. Although the prevalence of chronic lung disease in term-born CDH patients is around 30% and comparable with a preterm of around 1000 grams, no primary surfactant deficiency has been documented in newborns with CDH. The ongoing potential damage to the lungs by artificial ventilation may result in a secondary surfactant deficiency/inactivation. As such there is no argument to use artificial surfactant in patients with CDH and severe respiratory insufficiency. A recent trial [14] revealed that conventional artificial ventilation as initial ventilatory support is superior to high frequency oscillation. Increasingly treatment modalities are focusing on modulation of the pulmonary vascular tone as the main determinant of outcome in the majority of patients except for those who have an extreme form of pulmonary hypoplasia and the number of available alveoli is under a critical level. Our present level of knowledge of the molecular mechanisms underlying normal and abnormal pulmonary vascular development and the therapeutic targets to modulate pulmonary vascular growth and/ or tone is far from complete. As a consequence our treatment of pulmonary hypertension is still trial and error as no properly designed comparative effectiveness trials have been conducted so far for CDH specifically in contrast to the body of literature on the use of inhaled NO in preterm-born infants.
6.2.4Prenatal diagnosis
Worldwide, numerous attempts have been made to identify the optimal prenatal predictor of outcome for an individual patient with CDH. In most institutions the observed/expected (O/E) LHR (formerly LHR) which is independent of gestational age is used as a predictor of outcome of the individual patient although this use is still debated [15]. Apart from the O/E LHR, the position of the liver (liver up or down), the presence of associated anomalies (cardiac, renal, etc.) and the absence or presence of chromosomal abnormalities have all been used as predictors. It is important to understand that these values cannot be seen as independent to the standardization of postnatal care and experience of the treatment team. After years of compassionate use of fetoscopic endoluminal tracheal occlusion (FETO), to date the tracheal occlusion to accelerate lung growth (TOTAL) trial is on its way in both severe and moderate forms of prenatally diagnosed CDH guided by three European centers (Leuven, Belgium; London, UK; and Barcelona, Spain).
The results of this trial will determine the ultimate position of antenatal tracheal plugging in selected patients with prenatally diagnosed CDH. The results are expected in the next 2 years.
6.2.5Treatment
Following the ongoing data collection within the framework of the CDH, an international debate was started about optimal ventilatory support in newborns with CDH in need of artificial ventilation. In many institutions, high frequency oscillation (HFO) was used as a rescue therapy in case of ongoing high levels of arterial PCO2 which resulted in severe selection bias and the use of this treatment modality by unexperienced hands as a last resort. To answer the question about the optimal ventilatory approach in CDH, the so-called VICI-trial was conducted in prenatally diagnosed CDH patients, which revealed the superiority of conventional ventilation [13]. Against the background of international consensus on all treatment modalities in CDH, the results of this kind of investigations are of value and should be considered as a prerequisite of treatment results.
Long-term follow-up is equally important to evaluate the result of neonatal interventions with special attention to pulmonary, gastrointestinal and neurodevelopmental outcome parameters [16]. Evaluation of the presence of an increased pulmonary vascular tone by cardiac ultrasound is one of the contemporary hallmarks of therapy. Together with the evaluation of the presence of structural cardiac anomalies, the condition of the right ventricle is an important target of medical intervention for pulmonary vascular tone. In contrast to the positive effects of inhaled NO in preterms with pulmonary hypertension, the effects of inhaled NO in CDH newborns is around 30%.
Although no properly designed trials on the use of any vasoactive drug to modulate the pulmonary vascular tone have been conducted, most institutions use a combination of enhancing the systematic blood pressure by vasoactive drugs according to local habits in combination with inhaled NO. The medical therapy should be based on translational knowledge of the different pathways of endothelial smooth muscle cell interaction [17]. As part of the standard postnatal therapy, the ultimate aim is not to reach preductal situations of 100% in the absence of any R-L shunt because this will seriously damage the lungs due to increased shear forces and oxygen toxicity to the vulnerable lungs of CDH patients. In case all treatment modalities fail, extra corporeal membrane oxygenation (ECMO) can be used for specific patients and in institutions with a proven track record and enough patients on a yearly base. Solid guidelines for early transfer and clear criteria for admission to an ECMO center are important aspects of quality of care for the use of ECMO in CDH patients. Most institutions will still use VA-ECMO but from a patho-physiological point of view there are no arguments against the use of VV-ECMO in CDH patients as this will “oxygenate” the pulmonary vasculature and eventually result in relaxation of the abnormal pulmonary vascular bed. The application of ECMO has significantly decreased in many institutions over the years and ranges from 15–25% while the overall survival is around 35% post-ECMO according to the ELSO registry (2015).
6.2.6Surgical therapy
Timing and venue of surgery
A good outcome in pediatric surgery is not only dependent on surgical skills but even more on good collaboration between anaesthetists and surgeons. Especially for CDH this is particularly important. In the treatment of children born with CDH it was long thought that immediate closure of the defect in the diaphragm was the most important goal for both specialties. Since the 1980s, however, delayed repair of CDH became the accepted method, first allowing the infant to stabilize and obtain optimal pulmonary function prior to surgery [17]. Nevertheless, the definition of ‘stable’ was originally dependent on local expertise. A special sub group in this context is formed by premature babies with a CDH. Surgery within 48 hours in these infants is associated with a higher risk of intracerebral accidents. This phenomenon perhaps also holds true for instable term-born babies. The definition of early surgery in stable infants nowadays is still depending on local experiences and protocols and varies between 1 to 5 days. Until now there are no multi-center studies focussing on this issue [18].
In the 1990’s ECMO was incorporated in a strategy of delayed repair of CDH and was used for preoperative stabilization in patients who were unresponsive to maximal conventional treatment. If ECMO was required for preoperative stabilization, the diaphragmatic defect was repaired while the patient was on ECMO. In early experiences with this approach, all patients suffered from bleeding complications, however, since the introduction antifibrinolitic therapy with tranexamic acid (TEA) during and immediately after CDH repair on ECMO, hemorrhagic complications are significantly reduced [19]. Also the therapeutic or prophylactic use of antithrombotic agents such as TachoSil and Tissuecol is recommended. Most of the serious cardiopulmonary instable patients in ECMO centers will be placed on ECMO within 24 hours after birth. The timing of closure of the diaphragmatic defect in instable patients on ECMO is still being debated, however. Comparing surgery during the ECMO run with surgery after the ECMO run, further stabilization may be slightly favorable for the latter approach, at least in the CDH registry study [20]. The timing of surgery on ECMO is also being debated; either in the first 2 or 3 days on ECMO or at the expected end of the ECMO run. The advantage of early surgery is less edema but later surgery provides the opportunity to finish the ECMO run in case of bleeding.
The optimal venue for surgery is the operating room, which is a team-friendly site with enough space and availability of all facilities. However, surgery on these critically ill neonates in the intensive care unit does away with the potential risks involved in transport, such as temperature changes, metabolic de-arrangements and line or tube loss.
In addition, resuscitation procedures can be performed in an ideal environment and there will be continuity of care by the same physicians. No definitive data are available on the risk of infections.
Several papers have dealt with this issue with regard to various surgical disorders. There is only one study including CDH patients only [21]. The results of this study showed that patients operated on in the intensive care unit had a worse outcome as expressed by postoperative markers of inflammation and length of hospital stay. However, looking at the clinical characteristics, it appears that patients operated on in the intensive care unit were more severely ill than those operated on in the OR. In an attempt to exclude the impact of disease severity on the outcomes, a recent study focused on patients with a good postnatal prognosis defined as reaching stability within the first 72 hours of life [22]. In patients with high-risk CDH but a good postnatal outlook, the surgical venue does not have a significant impact on body temperature, infectious complications or respiratory outcomes.
Open surgical technique
In open surgery the most frequent incision is the subcostal one. It is essential that this incision is not too close to the ribs. Also a real transverse incision can be performed, which can give a better cosmetic result. Closing of the diaphragm depends on the size of the defect. The different phenotypes of CDH are well described by Ackerman [23]. In clinical practice and in the CDH registry the defect size is categorized into four groups: type A concerns small defects; type B are defects smaller than 50% of the diaphragm; type C defects are larger than 50% and type D concerns total agenesis [23]. Primary closure is mostly feasible for type A and small type B defects. In general this is performed with interrupted non-absorbable sutures. The larger type B, C and D defects need repair with a patch. The ideal patch material has to be strong enough to hold the sutures against abdominal pressure, pliable enough to allow natural movements and inert enough to prevent adhesions. Furthermore it would be ideal if native tissue could be incorporated to account for the rapidly expanding surface area preventing chest wall deformation. The most commonly used non-absorbable patch is made of polytetrafluorethylene (PTFE) but this does not allow incorporation of native tissue. In many small case studies absorbable patches are used, such as a-cellular porcine intestinal submucosa type 1 collagen (Surgisis), a-cellular human cadavaric dermis (AlloDerm), a-cellular porcine dermal cross-linked collagen (Permacol) or non-cross-linked collagen (Straties) [24]. In animal studies the PTFE patches show more recurrences, foreign body reactions and rib-deformation than most absorbable patches, however the long-term results in these animal models are not yet clear. Clinical data of small series with absorbable patches show recurrence rates between 14% and 40% and small bowel obstruction rates between 7% and 22%. Furthermore, these studies concern a relatively short follow-up while recurrences in the long term can also be expected due to impairment of the resulting scar tissue which is left at the place of the graft. In summary, none of the currently available patches hold all the ideal characteristics and the absorbable patches so far have not demonstrated clear advantages over the most used PTFE patch. The conventional open patch repair is with a patch the size of the defect or with an overlapping border of 1 cm circumferentially and sutured with interrupted non-absorbable material to the rim of the diaphragm. Significantly reduced recurrence rates in open repair are seen when cone-shaped PTFE patches are used [25]. The advantages of the cone-shaped patch are an increased abdominal capacity and reduction of redundant chest capacity, thereby allowing a normal physiological position of the abdominal organs, which in turn prevents gastroesophageal reflux (GER) and causes fewer recurrences because of separate fixation of the overlapping border of the cone. There are no studies comparing running and interrupted sutures; however, most centers use the latter technique. In addition to patch repair for large congenital diaphragmatic hernias a split abdominal wall muscle flap is a good alternative. A retrospective study showed a reduced risk of recurrence [26].
When closing the diaphragm, the intrathoracic abdominal organs have to be placed in the abdomen, which is sometimes too small for all these organs, in which case the abdominal cavity closing is impossible or results in abdominal compartment syndrome. In these cases, a patch is temporarily sutured in the abdominal wall. This can be performed with patches from PTFE, silicon or other materials. Normally these patches should be removed within a few weeks because of the high rate of infection. After this period, in most cases, the abdomen can be closed.
Originally thoracic drains were frequently used. It was important to have no suction on these drains, as this could move the mediastinum too much to the ipsilateral side. This can have a devastating effect on the circulatory and ventilatory stability of the patients. Today only the use of thoracic drainage is advocated in patients on ECMO.
Minimal access surgery
The traditional surgical management of CDH consists of repair by laparotomy. Since the 1990s minimal access surgery (MAS) became more popular in pediatric surgery. Therefore, laparoscopic and thoracoscopic repair techniques are being explored more in neonates with CDH. General advantages of MAS include less pain, less surgical stress, faster recovery and shorter hospitalization. Thoracoscopy for CDH can also bring a potential decrease in the occurrence of subsequent scoliosis, chest deformation, shoulder muscle girdle weakness and is associated with shorter duration of postoperative ventilation and a lesser need of narcotics. Early studies reported higher recurrence rates in the thoracoscopic repair group. This was explained by the so-called learning curve. However, more recent studies report no decrease of recurrence. Moreover, conversion of thoracoscopic surgery to open surgery is often needed (3.4% to 75.0%) because of surgical technical and ventilation problems. With respect to general complications, there is no significant difference between the approaches in the literature. In one study the mortality in the open repair group was higher due to these patients’ worse condition. Those patients have more need of cardiovascular and pulmonary support and often have a larger defect with the liver protruding into the thorax. All previous studies are retrospective and non-randomized. So the two types of surgery can still not be validly compared [27].
Several articles describe selection criteria for thoracoscopic repair. The cardiovascular criteria are almost the same in these overviews, i.e. no clinical signs of persistent pulmonary hypertension, no need for inhaled nitric oxide during surgery and no need for ECMO. They all agree that the patient has to be respiratory stable, but the definition of respiratory stable differs for PIP, PEEP, FiO2 and the saturation of peripheral oxygen (SpO2). The ventilation criteria of the recommendations of the CDH consortium are a PEEP of 2 to 5 cm and in addition FiO2 < 50% with SpO2 between 85%–95%, so that both open and thoracoscopic repair are possible. These recommendations are based on non-analytic studies, case reports or expert opinions. The artificial pneumothorax needed for thoracoscopic surgery creates acidosis due to hypercapnia by CO2 insufflation. This, in combination with higher intra-abdominal pressure, is believed to be related to deficient microcirculation. Nevertheless, patients with hypercapnia showed global hyperperfusion of the cerebral blood flow.
Bishay et al. recently proved severe arterial blood gas changes during thoracoscopic repair of CDH, but this finding was based on only five patients [28]. We also showed significant differences in pH and pCO2 values before and after thoracoscopic repair. However, these differences were small and have no clinical relevance. Due to the acidosis and hypercapnia, high frequency oscillation ventilation during neonatal thoracoscopic repair has gained attention in recent years. Mortellaro et al. showed that this allowed good intraoperative exposure in correction of esophageal atresia and CDH, while allowing excellent oxygenation and elimination of carbon dioxide to prevent acidosis [29].
The long-term consequences of hypercapnia and acidosis during surgery are still unknown.
Because of the possible detrimental effect in CDH patients of thoracoscopic repair, in several centers this is only done when the patient is cardiovascular and pulmonary stable. Still, if it appears that a patch is needed the technique must be converted to open surgery. The patient who does not fulfill the above criteria will undergo open repair from the start.
Other surgeries
The most common surgeries after recovering from the diaphragmatic hernia are fundoplications and gastrostomies. A high percentage of these patients have feeding problems which frequently require an open or laparoscopic gastrostomy. Also GER is seen in a high percentage of these patients, which leads to the need of a fundoplication [30].
6.2.7Long-term follow-up
As increasing numbers of neonates with CDH survive, attention should be directed to possibly persisting morbidity. Initially, most studies that had been published were cross-sectional and in small study populations and focusing on pulmonary or gastrointestinal morbidity. The past few years have seen a shift towards long-term multidisciplinary evaluations.
Respiratory morbidity
Despite the fact that severe respiratory failure with fulminant persistent pulmonary hypertension and the need for ECMO occurs in the neonatal period, airflow obstruction in survivors of school and adolescent age is usually mild [31, 32]. However, most studies were performed several decades ago and selection bias of the survivors may play a role as children with the most severe lung hypoplasia died. Longitudinal data in a cohort of ECMO-treated CDH patients showed that lung function seems to deteriorate over time as the children get older: mean (SE) z-score FEV1 decreased from ˗0.71(0.40) at 5 years to ˗2.73(0.61) at 12 years [33]. Airflow obstruction may occur from airway hypoplasia and fibrotic changes following lung damage caused by artificial ventilation and hyperoxia. Measured lung volumes are usually within normal limits. Most likely, the hypoplasia has been compensated for by alveolar distension and, consequently, hyperinflation. Therefore, the question arises whether measurement of lung function adequately reflects lung morphology. However, data on lung morphology are scarce and usually limited to postmortem findings. Recently, a new imaging technique has become available and the results of this technique in adult CDH survivors have been published: The use of hyperpolarized helium inhalation as contrast agent during magnetic resonance imaging (MRI) of the lungs provides the opportunity for non-invasive measurements of alveolar dimensions and evaluation of the homogeneity of ventilation (Fig. 6.2.1) [34].
Several cross-sectional studies on maximal exercise endurance at school age and adulthood are published: It is usually normal or decreased in comparison to healthy peers depending on the clinical characteristics of the study population and the reference data that have been applied [32, 35–37]. In a longitudinal study in CDH-patients treated with neonatal ECMO aged 5 to 12 years, maximal exercise endurance deteriorated significantly over time, the mean (SE) z-score maximal endurance time decreased from ˗0.53(0.33) at 5 years to ˗2.23(0.52) at 12 years [38]. It is not clear why exercise capacity decreases over time. No correlation between airflow obstruction and maximal exercise endurance has been observed. However, ventilation-perfusion mismatch, which has been described in CDH patients and gets worse when the children get older, may have a role in the deterioration of exercise capacity [39]. It is unknown whether pulmonary hypertension contributes to the deterioration of exercise capacity.
